Lismore RV-12

[aro]

or has it already happened?

It sounds similar to this incident:

 

http://www.vansairforce.com/community/showthread.php?t=59526

 

I think the cause in that case was determined to be the takeoff trim setting being incorrect. While the out of trim forces can be overcome, I suspect that taken by surprise, on a first flight, you may not be prepared to for the amount of forward stick force that is required.

 

 

[rgmwa]

So are we looking at an incorrectly wired elevator trim switch, or actuator, acting in reverse causing an uncontrolled pitch up ?.........Surley if that was the case, there would still be enough control authority to overcome that ?..........And my thoughts are if that can get past Wayne and the builder, with so many RV-12s out there is it going to happen again, or has it already happened?..............thoughts?....................Maj........

It appears that the trim was adjusted both up and down during the flight over a small proportion of its mapped -1 to +1 range. As Aro said, there have been a couple of incidents of stalls on initial liftoff caused by an incorrect initial setting of the trim position taking the pilot by surprise. Damage to the plane resulted but the pilots were OK. It has also been shown that the RV12 can be flown safely with the pitch trim at its extreme settings as might occur with a runaway trim motor.

 

rgmwa

 

 

[aro]

It appears that the trim was adjusted both up and down during the flight over a small proportion of its mapped -1 to +1 range.

I don't think the data presented tells us enough to know which direction the trim was moving, or through what percentage of its full travel.

 

** Speculation **

 

Possibly, the pilot recognised that the trim was incorrect and he was having to hold the nose down early in the takeoff roll. He held the nose down with forward pressure and tried to trim nose-down as the aircraft accelerated. The fluctuations in the pitch attitude at this time might suggest that the nose was trying to rise but was being held down.

 

Perhaps the trim was reversed, or perhaps the change in forces as the aircraft gained speed led the pilot to suspect it was reversed, so he tried trimming in the opposite direction.

 

Since the nose was being held down, the aircraft gained more speed than normal before liftoff.

 

In a tricycle, the centre of lift of the wing is ahead of the main wheels, so as the aircraft is allowed to lift off and weight transfers from wheels to wing the pitch up forces increase.

 

** More speculation **

 

The accident was the same cause as the other first flight stall on liftoff, but the extra speed from holding the node down more positively, and/or simply maintaining control for longer meant that the stall occurred at a higher altitude.

 

 

[Head in the clouds]

I just had another look at the graphs and I'm not sure whether we're interpreting them correctly or not. Some have talked about the ground roll and then sudden lift-off and pitch-up, others mention being airborne and then accelerating in ground effect and then the sudden pitch-up.

 

At first look I had assumed it was a pitch-up immediately following lift-off but on the second look I noticed the airspeed was in m/s not kts so when the pitch-up first starts the aircraft is doing 40m/s which is 144km/hr = 78kts.

 

So, based on that then the aircraft, being light, say half fuel, no bags, 1PoB probably got airborne at about 21.50:28 doing 16m/s = 31kt. There was slight pitch change starting 3 secs earlier which would most likely be the nosewheel lifting but not actually getting airborne. From 21.50:28 to 21.50:42 (14secs) the airspeed increases to 78kts but there is no pressure alt change indicating the acceleration is in ground effect and during that time the pitch trim is (apparently correctly) being applied to keep the nose from rising as might be expected during speed increase in ground effect.

 

From 21.50:41 the elevator trim moves quickly and smoothly in 3.5 secs from full forward to full aft and that seems very strange, whether the zoom climb was un-commanded or deliberate. If it was deliberate I don't think Wayne would perform it with a trim, in my experience he always flew with deliberation and overpowered control forces as required. When considering anything about whether the trim forces caught him unaware and/or produced heavy control forces let's not forget Wayne was a former contestant in the Mr Australia bodybuilding championships judged by Arnold Schwarzenegger no less, and was, at the time of this tragic event, a current fitness fanatic and active sports cyclist.

 

The pitch change from nose very high to nose very down, the stall, the rolling, the pressure alt change from max to descending all happen/begin at the same time (21.50:43) so to me that looks like a classic stall/spin entry off the top of the zoom climb, not the sort of mistake Wayne, a fully qualified and extremely active aerobatic flyer, would be likely to make. Note that during the dive the trim starts to wind forward again. If you were trying to fly the plane with the trim andthe nose was very down you'd be winding the trim back, not forward. So the trim's direction of action, and its indication, would appear to be correct for the take-off and acceleration but then incorrect after the pitch-up which says to me that the trim (or the indication of it) couldn't have been wired in reverse, it seems to have started out right and then gone wrong, and that indicates a failure I would think.

 

To follow the (logical?) sequence - Power on, lift nosewheel at 10m/s (= 19kts), airborne approx 2secs later at 31kts, accelerating to 78kts during the next 14secs at 20ft agl. Throughout those 14secs the trim is being increased for nose down (as you might expect) then suddenly the nose pitches up approx 60* which sounds exactly like the VAF link in aro's post #26 above. In the case described there the trim servo gears collapsed, so from that we can probably assume that there must be significant load on the trim servo. Later in that VAF thread (post #43) someone comments (as rgmwa mentions) that they went out and flew their RV12 with trim at both ends of the travel to check the control forces and it was quite flyable but the loads were significant and if you were caught by surprise it could be nasty. Also note that from the VAF thread the trim servo is supposed to be set to take 30-35 secs from one end of travel to the other, but in this event, after the pitch-up, the trim ran from full down indication to full up indication in 3.5secs, so was it set up 10 times too fast (unlikely I would think) or was it fully unloaded allowing it to run faster perhaps? Or was the indication not working correctly by that stage?

 

The trim servo indication is that the servo was running the other way once the nose pitched up, and that doesn't make sense, whether the pilot ran the servo or whether the EFIS has an 'auto' pitch feedback mechanism, either way the trim should have continued to try and run toward nose down once the nose was high, not trying to induce more nose up, and that is the main anomaly in the graph information as I see it. Maybe that means that any feedback info about the trim position comes from the trim servo rather than from the tab itself and so, if the trim servo collapsed the tab might have gone fully the other way in an instant causing the stick to leap back (perhaps out of the pilot's hand) and the trim information being fed back could have been anything at all if the gears were broken (think broken RC servo gears and the motor running constantly trying to equalise), accounting for the servo apparently running the wrong way.

 

It's just my amateur interpretation of course.

 

 

[aro]

Also note that from the VAF thread the trim servo is supposed to be set to take 30-35 secs from one end of travel to the other, but in this event, after the pitch-up, the trim ran from full down indication to full up indication in 3.5secs

We don't know the trim ran from one end of the travel to the other - in fact I think it didn't, based on the trim travel time you quote. We know that it ran one direction then the other from the starting position, that's all. The report says that the trim servo position seems to be mapped from -1 to +1. The data ranges from -0.005 to -0.03 which is a small fraction of that range. The reprot also states that further information is required to map those numbers to actual trim position.

 

 

[rgmwa]

From 21.50:41 the elevator trim moves quickly and smoothly in 3.5 secs from full forward to full aft and that seems very strange, whether the zoom climb was un-commanded or deliberate.

HIC, as aro suggests, "full forward to full aft" may not be correct. Quote from text in the report: "The elevator (stabilator) trim position appeared to have been mapped from -1 to +1; this range represents the travel of the motor arm attached to the pitch servo". If so, then the trim position value range on the graph (0 to -.03) suggests that only very small trim changes were made during the flight, and may in fact just be a reaction to aerodynamic loads or `slop' in the system. It's hard to say without more information, and we obviously need to be careful in trying to interpret the data. At the end of the day, perhaps the simplest explanation for an unrecoverable stall and wing drop at the top of a sharp low level climb may turn out to be the correct one.

 

rgmwa

 

 

[aro]

The increase in altitude happens in what looks like the first 3-5 seconds of the takeoff roll, when the airspeed is increasing between about 20 and 40 knots (converted from m/s). So I suspect the 20 foot altitude change is from the effect of slipstream over the static port, rather than accelerating in ground effect. I think during this time it is still on the runway.

 

 

[Head in the clouds]

HIC, as aro suggests, "full forward to full aft" may not be correct. Quote from text in the report: "The elevator (stabilator) trim position appeared to have been mapped from -1 to +1; this range represents the travel of the motor arm attached to the pitch servo". If so, then the trim position value range on the graph (0 to -.03) suggests that only very small trim changes were made during the flight, and may in fact just be a reaction to aerodynamic loads or `slop' in the system. It's hard to say without more information, and we obviously need to be careful in trying to interpret the data. At the end of the day, perhaps the simplest explanation for an unrecoverable stall and wing drop at the top of a sharp low level climb may turn out to be the correct one.rgmwa

The increase in altitude happens in what looks like the first 3-5 seconds of the takeoff roll, when the airspeed is increasing between about 20 and 40 knots (converted from m/s). So I suspect the 20 foot altitude change is from the effect of slipstream over the static port, rather than accelerating in ground effect. I think during this time it is still on the runway.

Yes, I see the point about the trim travel thing and most likely one would set the trim to neutral before take-off and then not worry about it again at least until the climb is established, and particularly not during the earliest stage of a first take-off in a new plane. The incident with the failed trim servo and very similar pitch-up on the other RV12 in 2010 described on the VAF thread perhaps concentrated my thinking on the trim factors.

 

aro - that's an interesting thought about about the static port, and might appear to be borne out by the absolutely flat pressure change graph until the pitch-up. But even on a runway it shouldn't be that flat perhaps, so maybe the pressure change sensor just wasn't sensitive enough. The thing that makes me tend away from your conclusion about it still being on the runway until the pitch-up is that Lismore runway is 54a (5400ft) and if he was still on the ground until the pitch-up he'd have had a ground-run (based on Google Earth and the painted runway markers) of about 1500ft or nearly 500m and be doing 78kts on the ground, and that doesn't seem likely. In a test flying situation (or perhaps any suitably cautious situation) I would have thought it would be more usual to lift off into ground effect, build safe speed and then climb away gently but something untoward appears to have taken place to cause that sudden pitch change.

 

 

[metalman]

Apart from trying to interpret the different graphs, I'm amazed at the amount of data available from a home built accident , it's getting up there with the commercial ops, barring a fire destroying stuff there's more info available to the bottom rung aviators than ever before

 

Matty

 

 

[crashley]

Have a look at the density altitude it appears that it was climbing at a rate of 2400 feet per minute or am I misreading the data

 

 

[Ultralights]

flying an aircraft with runaway trim is usually pretty easy, might be a bit tiring after a while, but easily doable. i have experienced it 2 times in aircraft with all flying stabilator. easily overcome. maybe we need to look at pilot incapacitation at some stage?

 

 

[rgmwa]

maybe we need to look at pilot incapacitation at some stage?

Not out of the question given the sudden pull-up. If you look at the data, the pitch change starts at about :42, but the altitude doesn't start to increase until :43.5 (lag due to inertia?) at which point the pitch angle also reaches its peak. The roll also commences at :43.5 (result of accelerated stall at low level due to pitch change with no altitude change?), and increases to 140 deg at the peak recorded pressure altitude of about 105 feet about 2 seconds later. It was all very quick.

 

rgmwa

 

 

[Head in the clouds]

Have a look at the density altitude it appears that it was climbing at a rate of 2400 feet per minute or am I misreading the data

Yes, the climb is rapid but the pressure sensor is obviously lagging behind by about 1.5secs as it shows the climb starting about 1.5secs after the pitch up but not the descent at all (which also took about 1.5secs).

 

crashley's observation about the rate of climb shows something else a bit strange. The pressure sensor shows a climb of 105ft in about 3 secs which is 2100ft/min and all of the sensors show that the climb lasted about 2.5-3 secs, so we should be able to safely assume that the aircraft only ever reached a height of approximately 105ft agl.

 

If we consider the plane was travelling at 78kts at the time of the pitch-up, if the engine was delivering full power during the climb then I would have thought the plane should have reached a much greater height before reaching the stall.

 

The flight profile of approx 60* climb angle starting from a speed of 78kts and only reaching a height of 105ft looks to me more like it was conducted without power, than with power ... I guess there's no engine data? At the funeral I did speak with some eyewitnesses who said the engine appeared to be running but it was too far away to tell if it was delivering any significant amount of power.

 

rgmwa just posted while I was writing this - note the 1.5secs without apparent climb is probably pressure sensor lag, just the same as our altimeter lags behind our climb.

 

 

[rgmwa]

rgmwa just posted while I was writing this - note the 1.5secs without apparent climb is probably pressure sensor lag, just the same as our altimeter lags behind our climb.

Could be HIC. Not sure why the roll would have commenced so early though. Anyway, enough speculation from me. I wasn't there, didn't see it, and don't know what caused the accident. Hopefully the coroner's report will pinpoint the cause.

 

rgmwa

 

 

[Ultralights]

78 kts, full power, sudden pull up, Possible G Stall situation?

 

 

[Powerin]

Just wondering, as I fly planes with stabilators.... A stabilator on its own induces very little stick force and so the trim tab is also an anti-servo tab to increase stick forces in pitch. Therefore a mechanical breakage of any of the linkage or trim motor would actually result in a lessening of stick forces. So a sudden increase in stick forces on a stabilator equipped aircraft would be caused by the trim motor itself physically pushing the tab in an unexpected direction. Is that correct?

 

 

[Guest Maj Millard]

Full- flying stabilators have to be very finely balanced statically, around their pivot point. The Partenavias is a good example of this, which is a copy of the Aztec set up. Don't know too much about the servo used in the RV-12, or just how much it actually does in the whole scheme of things.

 

Reading the trim servo notes on the Savannah the other day and I noticed the trim switch should be activated through its full range prior to flight..............Maj....

 

 

[turboplanner]

78 kts, full power, sudden pull up, Possible G Stall situation?

That seems to be the way the prune guys are heading, along with suggestions of reports from previous students of the same manoeuvre.

 

 

[Bandit12]

Unusual choice for a first flight though, and for someone who flys aeros, I would have expected more airspeed before a sudden pull up.

 

 

[Guest Maj Millard]

I don't seeing it as a intentional vertical pull- up at all, you just don't do that sort of thing on a first flight..even for Wayne. I expect it to be trim-actuator failure, or lack of static balance on the all- flying stabilator, that went full one way, and couldn't be moved. There has been one similar incident in the US..................Maj.....

 

 

[Guest Nobody]

There has been one similar incident in the US..................Maj.....

Was that one in an RV12 too? Do you know where and when? I had a look in the NTSB database but couldn't find it.

 

 

[rgmwa]

The RV-12 stabilator static balance is provided by lead weights fixed to the end of a long (about 600mm without going out to the shed to measure it) steel tube which is bolted to the front of the stab spar. The stab is actuated by cables fixed to horns above and below the spar. It would be difficult to overlook something critical to flight in a thorough pre-first flight inspection.

 

rgmwa

 

 

[planesmaker]

The RV-12 stabilator static balance is provided by lead weights fixed to the end of a long (about 600mm without going out to the shed to measure it) steel tube which is bolted to the front of the stab spar. The stab is actuated by cables fixed to horns above and below the spar. It would be difficult to overlook something critical to flight in a thorough pre-first flight inspection.rgmwa

rgmwa are you building an rv12? The anti servo tab on stabilator it would operate by deflecting more as stabilator is moved ie when up stabilator is used the trim tab would also go up thereby applying aerodynamic force to return stabilator to more neutral position yes? Is it at all possible (I would think it not possible)to assemble so that it becomes a servo tab ie it goes the opposite way thereby making stabilator wanting to go full travel either up or down(would be unstable)? Just trying to think what would cause such a strong pitch up. Tom

 

 

[rgmwa]

rgmwa are you building an rv12? The anti servo tab on stabilator it would operate by deflecting more as stabilator is moved ie when up stabilator is used the trim tab would also go up thereby applying aerodynamic force to return stabilator to more neutral position yes? Is it at all possible (I would think it not possible)to assemble so that it becomes a servo tab ie it goes the opposite way thereby making stabilator wanting to go full travel either up or down(would be unstable)? Just trying to think what would cause such a strong pitch up. Tom

Yes, almost finished. The trim tab (anti-servo tab more correctly) works the way you describe it, and it would be possible to get the servo wires reversed. However, it is not possible to get the arm that controls the tab wrong. It can only fit one way and will always move the tab in the right direction. The plans for the RV12 are exceptionally good, and the detailed `production acceptance procedures' checklist that comes with this kit would make it extremely unlikely that a servo wiring error would go unnoticed by the builder, the airworthiness inspector and the pilot. Also it takes around half a minute for the servo to drive the tab through its full range, so it could never generate a sudden pitch-up. The only things that I can imagine that would induce a sudden pitch-up are control cable failure or pilot action whether deliberate or involuntary.

 

rgmwa

 

 

[metalman]

There is another explanation to this accident, not nice or popular , but it is there